Michigan Maps
Group 4
April 22, 2020
## Warning: Removed 5 rows containing missing values (position_stack).
Figure Total accidents from 2016 to 2019 in each county in Michigan, higher populations typically result in more accidents.
Figure Total accidents (2016 - 2019) in each county of Michigan 
Figure Total population of each county of Michigan (2017 Estimate)
Figure Yearly accidents in each Michigan County from 2016 - 2020. 2016 and 2020 are misssing data.
Figure Decision tree splitting of Longitude and Latitude are shown as red lines on the map.
Figure Plot of Accidents per Capita in Michigan counties.
## Warning: Removed 1 rows containing non-finite values (stat_bin).
Figure Effect of temperature on accidents in Michigan
Figure More accidents occur during the day.
## # A tibble: 80 x 2
## County Accidents
## <fct> <int>
## 1 Alcona 5
## 2 Alger 1
## 3 Allegan 406
## 4 Alpena 20
## 5 Antrim 17
## 6 Arenac 35
## 7 Baraga 2
## 8 Barry 70
## 9 Bay 341
## 10 Benzie 28
## # … with 70 more rows## # A tibble: 80 x 2
## County Accidents
## <fct> <int>
## 1 Genesee 27072
## 2 Wayne 22214
## 3 Kent 12989
## 4 Oakland 10115
## 5 Macomb 4361
## 6 Washtenaw 1689
## 7 Ingham 1126
## 8 Ottawa 941
## 9 Livingston 723
## 10 Jackson 709
## # … with 70 more rowsTable Total Accidents from 2016 - 2020 in each county.
##
## Call:
## lm(formula = Accidents ~ Sunrise_Sunset + County, data = train_matct)
##
## Residuals:
## Min 1Q Median 3Q Max
## -214.680 -48.827 -5.781 51.923 290.430
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 397.570 12.584 31.593 < 2e-16 ***
## Sunrise_SunsetNight -120.890 8.947 -13.512 < 2e-16 ***
## CountyIngham -326.123 16.815 -19.394 < 2e-16 ***
## CountyKent -179.000 16.634 -10.761 < 2e-16 ***
## CountyMacomb -284.833 16.634 -17.123 < 2e-16 ***
## CountyOakland -213.125 16.634 -12.812 < 2e-16 ***
## CountyWashtenaw -320.470 16.815 -19.058 < 2e-16 ***
## CountyWayne -62.562 16.634 -3.761 0.000201 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 81.49 on 324 degrees of freedom
## Multiple R-squared: 0.7276, Adjusted R-squared: 0.7218
## F-statistic: 123.7 on 7 and 324 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = Accidents ~ Sunrise_Sunset + County + Month, data = train_matct)
##
## Residuals:
## Min 1Q Median 3Q Max
## -206.679 -52.052 -6.275 50.798 267.871
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 434.057 19.251 22.547 < 2e-16 ***
## Sunrise_SunsetNight -121.129 8.879 -13.643 < 2e-16 ***
## CountyIngham -326.792 16.688 -19.583 < 2e-16 ***
## CountyKent -179.000 16.502 -10.847 < 2e-16 ***
## CountyMacomb -284.833 16.502 -17.260 < 2e-16 ***
## CountyOakland -213.125 16.502 -12.915 < 2e-16 ***
## CountyWashtenaw -320.638 16.688 -19.214 < 2e-16 ***
## CountyWayne -62.562 16.502 -3.791 0.00018 ***
## Month02 -27.393 21.606 -1.268 0.20581
## Month03 -13.929 21.606 -0.645 0.51963
## Month04 -40.857 21.606 -1.891 0.05955 .
## Month05 -41.214 22.028 -1.871 0.06227 .
## Month06 -38.250 21.606 -1.770 0.07765 .
## Month07 -62.069 21.810 -2.846 0.00472 **
## Month08 -38.989 21.810 -1.788 0.07480 .
## Month09 -36.893 21.606 -1.707 0.08872 .
## Month10 -32.893 21.606 -1.522 0.12893
## Month11 -70.536 21.606 -3.265 0.00122 **
## Month12 -33.393 21.606 -1.546 0.12323
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 80.84 on 313 degrees of freedom
## Multiple R-squared: 0.7411, Adjusted R-squared: 0.7262
## F-statistic: 49.76 on 18 and 313 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = Accidents ~ Sunrise_Sunset + County + Month, data = train_matct)
##
## Residuals:
## Min 1Q Median 3Q Max
## -206.679 -52.052 -6.275 50.798 267.871
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 434.057 19.251 22.547 < 2e-16 ***
## Sunrise_SunsetNight -121.129 8.879 -13.643 < 2e-16 ***
## CountyIngham -326.792 16.688 -19.583 < 2e-16 ***
## CountyKent -179.000 16.502 -10.847 < 2e-16 ***
## CountyMacomb -284.833 16.502 -17.260 < 2e-16 ***
## CountyOakland -213.125 16.502 -12.915 < 2e-16 ***
## CountyWashtenaw -320.638 16.688 -19.214 < 2e-16 ***
## CountyWayne -62.562 16.502 -3.791 0.00018 ***
## Month02 -27.393 21.606 -1.268 0.20581
## Month03 -13.929 21.606 -0.645 0.51963
## Month04 -40.857 21.606 -1.891 0.05955 .
## Month05 -41.214 22.028 -1.871 0.06227 .
## Month06 -38.250 21.606 -1.770 0.07765 .
## Month07 -62.069 21.810 -2.846 0.00472 **
## Month08 -38.989 21.810 -1.788 0.07480 .
## Month09 -36.893 21.606 -1.707 0.08872 .
## Month10 -32.893 21.606 -1.522 0.12893
## Month11 -70.536 21.606 -3.265 0.00122 **
## Month12 -33.393 21.606 -1.546 0.12323
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 80.84 on 313 degrees of freedom
## Multiple R-squared: 0.7411, Adjusted R-squared: 0.7262
## F-statistic: 49.76 on 18 and 313 DF, p-value: < 2.2e-16Figure Summary of linear regression models. The first linear regression model with predicting accident volume by county and day vs night, Test MSE is lower than second model and is used to generate the figures below.
Figure Using linear regression to predict the number of daytime accidents in each month of the year in Wayne County.
Figure Using linear regression to predict the number of nighttime accidents in each month of the year in Wayne County.
Figure Using linear regression to predict the number of daytime accidents in each month of the year in Genesee County.
Figure Using linear regression to predict the number of nighttime accidents in each month of the year in Genesee County.
Figure Using linear regression to predict the number of daytime accidents in each month of the year in Kent County.
## fit lwr upr
## 1 97.68016 72.92307 122.4373
Figure Using linear regression to predict the number of nighttime accidents in each month of the year in Kent County.
## Loading required package: carData##
## Attaching package: 'car'## The following object is masked from 'package:purrr':
##
## some## The following object is masked from 'package:dplyr':
##
## recode## Anova Table (Type III tests)
##
## Response: Accidents
## Sum Sq Df F value Pr(>F)
## (Intercept) 6421246 1 966.94 < 2.2e-16 ***
## Sunrise_Sunset 1212353 1 182.56 < 2.2e-16 ***
## County 4596049 6 115.35 < 2.2e-16 ***
## Residuals 2151614 324
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1Figure Anova analysis shows that more accidents occur during the day.
| City | Accidents |
|---|---|
| Flint | 13533 |
| Detroit | 13459 |
| Grand Rapids | 11913 |
Table Three cities (Flint, Detroit, and Grand Rapids) in Michigan account for the majority of the accidents reported in the state from 2016-2019.
## Source : https://maps.googleapis.com/maps/api/staticmap?center=wayne%20county&zoom=10&size=640x640&scale=2&maptype=terrain&language=en-EN&key=xxx-ByU## Source : https://maps.googleapis.com/maps/api/geocode/json?address=wayne+county&key=xxx-ByU## Warning: Removed 1 rows containing missing values (geom_point).
Figure Accidents in Wayne County from 2016 - 2019.
## Source : https://maps.googleapis.com/maps/api/staticmap?center=flint&zoom=12&size=640x640&scale=2&maptype=terrain&language=en-EN&key=xxx-ByU## Source : https://maps.googleapis.com/maps/api/geocode/json?address=flint&key=xxx-ByU## Warning: Removed 10855 rows containing missing values (geom_point).
Figure Accidents near Flint.
## Source : https://maps.googleapis.com/maps/api/staticmap?center=I-75,%20M118,%20Flint&zoom=14&size=640x640&scale=2&maptype=terrain&language=en-EN&key=xxx-ByU## Source : https://maps.googleapis.com/maps/api/geocode/json?address=I-75,+M118,+Flint&key=xxx-ByU## Warning: Removed 24498 rows containing missing values (geom_point).
Figure Accidents near I-75 Exit 118.
## Source : https://maps.googleapis.com/maps/api/staticmap?center=grand%20rapids&zoom=12&size=640x640&scale=2&maptype=terrain&language=en-EN&key=xxx-ByU## Source : https://maps.googleapis.com/maps/api/geocode/json?address=grand+rapids&key=xxx-ByU## Warning: Removed 1089 rows containing missing values (geom_point).
Figure Accidents near Grand Rapids.
Figure Total number of accidents in the dataset for each season.
Figure Count of severity at each level (1-4) for each month. Severity does not differ significantly month to month.
Figure Correlation Plot of Accidents and Accident per capita vs. all other census variables.
Figure Included Severity Per Captita which positively correlates to the Income Per Capita census variable.
## Source : https://maps.googleapis.com/maps/api/staticmap?center=michigan&zoom=6&size=640x640&scale=2&maptype=terrain&language=en-EN&key=xxx-ByU## Source : https://maps.googleapis.com/maps/api/geocode/json?address=michigan&key=xxx-ByU
Figure Density plot of accidents in Michigan.